| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH COMMUNICATION |
1 From the Departments of Ophthalmology and Visual Sciences (JAM, TLL, and SMM) and of Biostatistics and Medical Informatics (MJG and RJC), University of Wisconsin, Madison, WI; the Department of Ophthalmology, Medical College of Georgia, Augusta, GA (DMS); the Casey Eye Institute, Oregon Health Sciences University, Portland, OR (MLK); the Department of Psychology, Brown University, Providence, RI (BRW); and the Jean Mayer, US Department of Agriculture Human Nutrition Research Center on Aging at Tufts University, Boston, MA (EJJ)
2 This research was part of the Carotenoids and Age-Related Eye Disease Study (CAREDS), an ancillary study of the Women's Health Initiative. 3 Supported by NIH grant EY13018 and DK 07665. The National Eye Institute and Research to Prevent Blindness funded CAREDS, and the National Heart, Lung and Blood Institute of the National Institutes of Health, US Department of Health and Human Services funded the WHI program. 4 Reprints not available. Address correspondence to JA Mares, University of Wisconsin, Department of Ophthalmology and Visual Sciences, 610 North Walnut Street, 1063 WARF Building, Madison, WI 53726-2336. E-mail: jmarespe{at}facstaff.wisc.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Objective: We aimed to investigate the predictors of the density of lutein and zeaxanthin in the macula of the retina.
Design: Macular pigment optical density (MPOD) was measured by heterochromatic flicker photometry. Relations to dietary lutein and zeaxanthin and to other predictors were measured in 1698 women aged 53–86 y. The women were members of observational study cohorts of the Women's Health Initiative at Iowa City, IA, Madison, WI, or Portland, OR, and participated in the Carotenoids in Age-Related Eye Disease Study (2001–2004).
Results: MPOD at 0.5 degrees from the foveal center was 30% higher in women in the highest quintile for lutein and zeaxanthin intake [
(±SD): 0.40 ± 0.21] than in women in the lowest quintile (0.31 ± 0.21) and 20% higher after adjustment for other predictors. Dietary intake of lutein, zeaxanthin, fiber, and polyunsaturated fatty acids (% of energy) together explained 3% of the variability in MPOD. Higher waist circumference and diabetes, which are related to lower MPOD, together with study site explained an additional 5% of variation. The total explained variability increased to 12% when lutein and zexanthin concentrations obtained from the serum, which were collected 4-7 y earlier, were added to the model.
Conclusions: MPOD is directly related to dietary intake of lutein and zeaxanthin but even more strongly to serum concentrations, which may reflect unmeasured physical and medical factors that influence the uptake, distribution, and utilization of lutein and zeaxanthin. Higher abdominal body fat and diabetes are related to lower MPOD. Unknown predictors of retinal carotenoids remain.
Key Words: Lutein • zeaxanthin • carotenoids • retina • blood • diet
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The only source of these plant pigments in the eye is diet. Current studies indicate an ability to influence concentrations of lutein and zeaxanthin in macular pigment by increasing concentrations taken in from foods (7-9) or supplements (10-12). However, there is a variable response of retinal carotenoids to supplementation (8, 12-14). The degree to which variations in dietary lutein and zeaxanthin relate to tissue concentrations needs to be understood to make recommendations regarding appropriate concentrations of lutein and zeaxanthin in the diet or supplements.
The potential for these carotenoids to protect against age-related eye diseases is supported by their ability to reduce oxidative stress, absorb blue light, and stabilize cell membranes [reviewed in Krinsky et al (15)]. To date, several small studies (9–63 persons) suggest that people with AMD (16, 17) or high lens density (18, 19) have lower concentrations of lutein and zeaxanthin in the macula than do people without AMD or high lens density. However, one other study conducted in 435 persons did not show this relation (20). Epidemiologic studies are needed to assess the influence of these carotenoids on the risk of cataract and AMD over the long term and under various environmental and genetic exposures. A need exists to identify the influences of macular pigment optical density (MPOD) so they can be measured and adjusted for in epidemiologic studies.
Previous studies in samples of 88 to 376 persons have provided clues about dietary, physical, and lifestyle factors that influence concentrations of carotenoids in the retina. Light eye color (21-23), smoking (24), high levels of body fat (7, 13, 25), and female sex (21-23) have been related to lower MPOD in some, but not all, study samples. Blood concentrations of lutein and zeaxanthin are associated with endogenous hormone status (26), alcohol use (27), and level of physical activity (28). These may influence concentrations of lutein and zeaxanthin in the retina as well but have not been investigated. Because of the many factors involved, the predictors of macular pigment need to be investigated in large samples of people, with adjustment for all suspected predictors of macular pigment identified to date.
The Carotenoids in Age-Related Eye Disease Study (CAREDS), an ancillary study of the Women's Health Initiative, was designed to investigate the relations of dietary lutein and zeaxanthin to these eye diseases and to extend previous small studies of predictors of MPOD. We report the degree to which concentrations of lutein and zeaxanthin in the diet are related to MPOD. We also report the other dietary, serum, lifestyle, physical, and medical attributes that are associated with MPOD in this sample.
|
SUBJECTS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
80th and 
20th percentiles for lutein and zeaxanthin intakes at WHI baseline. After enrolling slightly more than one-half of the participants, we observed response rates of 74% and 65% in women in the high and low quintiles of lutein intake, respectively. The expansion of the recruitment pool, by using wider percentile cutoffs, was adopted to achieve needed sample sizes and approximately equal numbers of women with high and low intakes. Women at these extreme levels of lutein intake, combined, did not differ significantly in age, education, body mass index, smoking status, use of supplements or hormone replacement therapy, or history of diabetes or cardiovascular disease from those with intakes between the 28th and 78th percentiles (data not shown).
The original cohort was recruited to the WHI in each of these regions through regional mass-mailings and mass-media strategies. The WHI is a prospective cohort study of the most common causes of mortality and morbidity in 
93 676 postmenopausal women aged 50–79 y from 40 sites across the United States. Women were excluded if they had medical conditions that predicted survival of <3 y, alcoholism, drug dependency, or mental illness.
Of the 3140 women who were selected based on lutein and zeaxanthin intake at WHI baseline study visits, 93 women died or were lost to follow-up between selection in year 2000 and enrollment in CAREDS from 2001 to 2004. The remaining women were sent letters inviting them to participate in CAREDS. A total of 1042 women declined participation and 2005 were enrolled. The participants had diets that were higher in lutein and zeaxanthin (median: 1030 compared with 952 mg/d) and lower in fat (31% compared with 33% of energy) than those who declined and were more likely to use nutrient supplements regularly (60% compared with 55%) and hormone replacement therapy (54 compared with 43%). Compared with nonparticipants, the participants were more likely to have graduated high school (77% compared with 66%) and less likely to smoke (4% compared with 7%) or have diabetes (4% compared with 6%). Median waist circumference was lower in participants than in non participants (82 compared wtih 86 cm.). Of these women, 1894 participated in study visits and 1803 completed MPOD measurements. A total of 105 women were excluded from the analysis dataset because of missing data on covariates. Thus, 1698 women comprised the analysis dataset for the present investigation. This sample of women aged between 53 and 86 y were 98% white and highly educated (78% had some college education). All procedures conformed to the Declaration of Helsinki and were approved by the Institutional Review Board at each University.
Dietary lutein and zeaxanthin estimates
Diet was assessed at WHI baseline (1994–1998; as a selection criteria) and CAREDS baseline (2001–2004; to study in relation to MPOD) by using a semiquantitative food-frequency questionnaire that was based on the instrument used in the Women's Health Trial Feasibility Study (29). Nutrient estimates in WHI and CAREDS were computed from responses to food-frequency questionnaires at the Fred Hutchinson Cancer Research Center. Estimates of lutein and zeaxanthin intake at both time periods were modified by responses to a query about the proportion of time that dark compared with light greens were consumed in salads at the time of CAREDS and in an earlier time period (1986–1988). Average daily intake of lutein and zeaxanthin from supplements was computed from responses to vitamin supplement questionnaires collected at CAREDS study visits that queried for the dose, frequency, and duration of specific supplement intake. Intake from food and supplemental sources were summed in analyses.
Serum analyses of lutein and zeaxanthin
Serum samples, which were obtained from participants in WHI baseline examinations (1994–1998) and stored at –80 °C, were analyzed for concentrations of trans-lutein and -zeaxanthin at Tufts University by a reverse phase HPLC analysis (30) to evaluate relations to the prevalence of eye diseases in CAREDS (31). In the present study, we describe relations of serum lutein and zeaxanthin to MPOD in CAREDS. Blind duplicates were analyzed in each batch of serum analyses (in a total of 57 participants). Mean CVs were 7.0% for all trans-lutein and 9.6% for all trans-zeaxanthin. Total triacylglycerols and cholesterol were measured by an automated chemistry analyzer based on the cholesterol oxidase method (32, 33).
MPOD measurements
Measurements were made by using a standardized protocol by the psychophysical method of heterochromatic flicker photometry during CAREDS study visits conducted from 2001to 2004. This protocol, which was described in detail previously, had high test-retest reliability (r = 0.9) and participant responses at 2 wavelengths that are consistent with the absorption spectrum of lutein and zeaxanthin (34). Briefly, the participants were fitted with trial frames and appropriate lenses for testing after refraction. The best flicker frequency for heterochromatic flicker photometry was then established for each participant. The participant made 4 separate determinations for each target at several retinal locations. The participant viewed a small test field superimposed on a blue background with the right eye. The test field alternates between a wavelength (blue or blue-green) that is absorbed by the macular pigment (MP) and a reference wavelength (green to yellow-green) that is outside the absorption band of MP. When the frequency of alternation is chosen correctly, the test field appears to flicker. When making measurements, the participant is instructed to adjust the energy of the bluish test light so that the flicker stops. The amount of bluish light that is required to nullify the flicker provides a measure of the absorption of MP (ie, MPOD) at the retinal location of the test light. The participant was instructed to fixate at the center of the following targets: 0.25, 0.50, 1.00, and 1.75 degrees from the foveal center so that MPOD at different eccentricities as measured in reference to a target at 7 degrees from the center. Measurements were also made with the left eye at 0.5 and 1.0 degree from the foveal center.
Covariates
During CAREDS study visits, weight (with a calibrated scale), height (with a stadiometer), and waist and hip circumferences (with a tape measurement 1 inch above navel and at widest spot between the waist and hips) were measured. The presence of physician-diagnosed diabetes, family history of cataract (immediate family member diagnosed before the age of 65 y) and macular degeneration (in an immediate family member aged 55 y), use of cholesterol-lowering medications, and ambient sunlight exposure (in the past 20 y; based on outdoor activities during routine and vacation periods, living location, and use of protective gear, ie, hats and sunglasses) was obtained from questionnaires answered by the participants at CAREDS study visits. Some demographic and lifestyle data (education, smoking status, physical activity, use of hormone replacement therapy, and alcohol use) were available from questionnaires completed at WHI study entry and updated in questionnaires at CAREDS study visits (smoking status and use of hormone replacement therapy and alcohol).
Statistical analyses
Mean MPOD levels and spatial distribution were computed by quintile of lutein and zeaxanthin in diets at CAREDS baseline (2001–2004). The correlations between MPOD and other dietary, lifestyle, serum, physical, and medical characteristics were determined by computing Pearson correlation coefficients before and after adjustment for dietary lutein and zeaxanthin.
Mean MPOD was computed by quintile of dietary lutein and zeaxanthin and adjusted for other dietary characteristics that were related to MPOD by using general linear regression modeling. The covariates that showed the largest MPOD differences across categories were entered first in the model, and other covariates were tested subsequently in an iterative process. Those that were significantly related to MPOD (P < 0.05) were retained in the adjusted model. Next, the relation of diet lutein and zeaxanthin to nondietary correlates was measured in the same way. Variables in both models were finally tested in a combined regression model. To gain insight into which correlates of MPOD may reflect the ability to absorb and retain carotenoids in the blood, we investigated the correlates of serum lutein and zeaxanthin in a similar way. When relations of diet lutein and zeaxanthin to MPOD differed by measured variables (P < 0.1), we computed the relation of dietary lutein and zeaxanthin in subgroups for which an interaction was detected.
Finally, we made an a priori plan to explore possible predictors of poor MPOD response to high dietary lutein and zeaxanthin by cross-classifying women based on quartiles of lutein and zeaxanthin in serum and diet at WHI baseline and comparing the distribution of covariates. We confined these analyses to women with dietary intakes above the median, so that low serum carotenoid concentrations did not reflect low dietary carotenoids. Low responders were defined as having serum lutein and zeaxanthin in the lowest quartile and diet lutein and zeaxanthin intake in the 2 upper quartiles or having serum lutein and zeaxanthin in the second quartile and lutein and zeaxanthin intake in the highest quartile. High responders were defined as having dietary lutein and zeaxanthin in the highest quartile and serum concentrations in the third or fourth quartiles. All analyses were performed with the use of SAS software, version 8.12 (SAS Institute Inc, Cary, NC).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. In women with intakes of lutein and zeaxanthin in the high and low quintiles, MPOD declined sharply from 0.25 degrees to 1.75 degrees from the foveal center. In the entire sample, mean (±SD) MPOD was 0.43 ± 0.23 at 0.25 degrees and 0.13 ± 0.16 at 1.75 degrees in the right eye. MPOD in right eyes were slightly higher than in left eyes. As measured in the present study, mean (±SD) MPOD in right and left eyes was 0.36 ± 0.22 compared with 0.34 ± 0.21 (P < 0.0001) and 0.26 ±0.19 compared with 0.24 ± 0.18 (P < 0.0001) at 0.5 and 1.0 degrees from foveal center, respectively. Because the predictors of MPOD did not differ significantly by the distance from the foveal center (not shown), subsequent data are presented for predictors of MPOD at 0.5 degrees only, the eccentricity with the lowest within participant to between participant variability (34).
|
and Figure 2) was 31% higher in the fifth than in the first quintiles of dietary lutein and zeaxanthin. When lutein and zeaxanthin were expressed as a nutrient density (in µg/1000 kcal), the average difference between quintile 1 and 5 was identical (not shown.) Overall, MPOD was correlated with dietary lutein and zeaxanthin intake (r = 0.16, P < 0.0001). Expressing lutein and zeaxanthin intake per 1000 kcals lowered the correlation with MPOD to 0.06. Therefore, in remaining analyses, dietary lutein and zeaxanthin intake was not adjusted for energy intake. After adjustment for other correlates of MPOD (Table 1), the relation of dietary lutein and zeaxanthin to MPOD was slightly attenuated such that MPOD at 0.5 degrees from the foveal center was 20% higher in fifth than in the first quintile of dietary lutein and zeaxanthin (Figure 2). A small increase in mean MPOD was observed beyond the third quartile of lutein and zeaxanthin in the overall sample and in subgroups stratified by age, education, and all other ocular, medical, and lifestyle factors (not shown) with the following exceptions: the nature of the relation of dietary lutein and zeaxanthin to MPOD differed in women with diabetes (Figure 3) and, nonsignificantly (P = 0.13), in women in extreme quartiles for waist circumference. Women in the highest quartile for waist circumference had mean (±SE) MPODs of 0.37 ± 0.03 and 0.47 ± 0.04 for the fourth and fifth quintiles for dietary intake of lutein and zeaxanthin, respectively. Women in the lowest quartile for waist circumference had mean (±SE) MPODs of 0.24 ± 0.07 and 0.38 ± 0.02 for the fourth and fifth lutein and zeaxanthin intake quintiles, respectively.
|
|
|
). Although BMI and the waist-to-hip ratio were also correlated with MPOD, they were not entered separately into adjusted models because of their high correlation with waist circumference (r = 0.86 and 0.71, respectively). Across all quintiles of dietary lutein and zeaxanthin, MPOD was lower in the women in the highest quartile for waist circumference than in the women in the lowest quartile (not shown).
Diabetes
Diabetes was the next most strongly associated variable with MPOD. After adjustment for all other correlates, having diabetes was associated with 19% lower MPOD compared with not having diabetes (Table 1). Diabetes also modified the relation of dietary lutein and zeaxanthin to retinal carotenoid levels: in women with diabetes, MPOD was 200% higher in the highest compared with the lowest quintile of lutein and zeaxanthin intake, compared with only 28% higher in women without diabetes (P for interaction = 0.05) (Figure 3).
Other predictors
Other variables that were significantly associated with MPOD included study site, level of education, physical activity, and a family history of AMD. After adjustment for the concentration of lutein and zeaxanthin in the diet, only study site remained significantly related to MPOD (Table 1). Mean MPOD was 15% lower in women from Wisconsin than in women from Iowa and Oregon, even after adjustment for all other measured correlates of MPOD despite no significant differences in any demographic, physical, lifestyle, dietary, or health variables measured across the women from these 3 samples (not shown). Some attributes that have been associated with MPOD in previous studies (eg, age and smoking status) were unrelated to MPOD in the current sample before or after adjustment for other correlates. Other medical factors that were not associated with MPOD in this sample (and are excluded for brevity) are presence of AMD (LaRowe et al, unpublished observations) and cataract (Moeller et al, unpublished observations) and history of cardiovascular disease (data not shown).
After adjustment for lutein and zeaxanthin in the diet and for all other nondietary correlates, some aspects of diet remained independently associated with MPOD: MPOD was higher in women with higher intakes of fiber and polyunsaturated fatty acids (% of energy) (Table 1). High dietary intakes of ß-carotene, fiber, fruits and vegetables, and n–3 fatty acids were associated with high MPOD. However, these associations were no longer significant after adjustment for dietary levels of lutein and zeaxanthin (Table 1).
Relation of serum lutein and zeaxanthin to MPOD
Concentrations of lutein and zeaxanthin in serum, which was collected 4–7 y earlier, were correlated (Pearson correlation coefficient r = 0.31) with MPOD. MPOD in women in the highest quintile of serum lutein and zeaxanthin was almost 2-fold that in women in the lowest quintile (Table 1). Adjustment for serum cholesterol, which itself was correlated with serum lutein and zeaxanthin (r = 0.17), did not increase the correlation between serum lutein and zeaxanthin and MPOD. Serum concentrations of lutein and zeaxanthin individually, whether adjusted (r = 0.32 and 0.24, respectively) or unadjusted (Pearson correlation coefficient r = 0.31 and 0.23, respectively) for serum cholesterol, did not improve the correlation with MPOD. For this reason, all subsequent investigations of serum lutein and zeaxanthin used the combined carotenoids unadjusted for serum cholesterol.
Predictors of serum lutein and zeaxanthin
We investigated the dietary, lifestyle, demographic, physical, and medical predictors of serum lutein and zeaxanthin to better understand the degree to which potential predictors of MPOD reflect their relation to levels of lutein and zeaxanthin in serum in this sample (Table 2). The most important predictor of serum lutein and zeaxanthin at WHI baseline was the amount of dietary lutein and zeaxanthin intake (r = 0.39). Serum lutein and zeaxanthin increased by 64% between the first and fifth quintiles for dietary intake of lutein and zeaxanthin. Serum lutein and zeaxanthin were also related to the intake of fruits and vegetables, generally (r = 0.36), and some substances they contain, incluidng dietary ß-carotene (r = 0.34) and fiber (r = 0.31), but these relations were strongly attenuated after adjustment for dietary lutein and zeaxanthin. Concentrations of total fat (% of energy) and n–3 fats in the diet were inversely related to serum concentrations of lutein and zeaxanthin.
|
-carotene). This was true even after adjustment for dietary levels of lutein and zeaxanthin and additional adjustment for fruit and vegetable intake. After adjustment for dietary lutein and zeaxanthin, serum lutein and zeaxanthin were directly associated with study site and levels of physical activity, serum cholesterol, alcohol intake, and use of hormone replacement therapy. Serum lutein and zeaxanthin were inversely associated with waist circumference, BMI, waist-to-hip ratio, serum triacylglycerols, diabetes and use of cholesterol-lowering medications. Of these nondietary attributes related to serum lutein and zeaxanthin, waist circumference was the most strongly correlated variable (r = –0.34). Diabetes and serum triacylglycerols were no longer significantly correlated with serum lutein and zeaxanthin after adjustment for waist circumference. Family history of cataract or AMD, iris color, cigarette smoking, and sunlight exposure were not significantly related to serum concentrations of lutein and zeaxanthin (not shown).
Variation in MPOD explained by measured factors
In the fully adjusted model, we evaluated the degree to which factors measured in the current study explain variability in MPOD. Overall, only 8% of the variation in MPOD could be explained by the dietary, lifestyle, medical, and physical variables measured in CAREDS. Only 3% of the variation in MPOD was explained by the intake of dietary lutein, zeaxanthin, fiber, and polyunsaturated fat (% of energy). Five percent of the variation in MPOD was explained by nondietary variables that were independently related to MPOD (ie, waist circumference, diabetes, and study site). The total explained variability in MPOD increased from 8% to 12% when lutein and zeaxanthin concentrations obtained from the serum collected 4-7 y earlier were added to the model.
Possible low responders to high dietary lutein and zeaxanthin intake
Finally, we compared women with relatively high intakes of lutein and zeaxanthin above the median who also had corresponding high and low concentrations of lutein and zeaxanthin in the serum in an attempt to identify characteristics that may reflect low absorption, high turnover, or low serum or tissue distributions of lutein and zeaxanthin. Low responders had higher waist circumferences, serum triacylglycerol, and levels of dietary fat (% of energy) than did high responders and were more likely than high responders to have come from the Iowa study site, have low levels of physical activity or alcohol intake, have taken cholesterol-lowering medications, or have diabetes (Table 3). No significant differences in family history of AMD, iris pigmentation, history of smoking, or use of hormone replacement therapy were observed between the groups (not shown). Women with possible low serum response had lower mean (±SD) MPOD (0.29 ± 0.20) than did women with high serum response (0.44 ± 0.23).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2.2 mg lutein + zeaxanthin/d (Figure 2). However, these data are insufficient to conclude that there is no benefit in intakes of lutein and zeaxanthin beyond 2–3 mg/d. This is because the nature of this relation, although stable across many sample strata (not shown), differs in some subgroups: in women with diabetes (Figure 3) and in women in high or low quartiles for waist circumference (see Results), MPOD means do not plateau at lutein and zeaxanthin intakes in the range of 2 mg/d as they do in the overall sample. Experimental studies are needed to directly establish the influence of dietary lutein and zeaxanthin on their concentrations in the retina. Moreover, the individual MPOD responses to supplementation with xanthophylls are highly variable across persons. Between 20% and 50% of participants in previous investigations have shown low serum response, low retinal response, or both to oral supplementation with these carotenoids (8, 11, 12 ). In general, blood responses to oral lutein and zeaxanthin carotenoids are variable across people (35, 36).
Estimates of body fat and diabetes
Higher waist circumferences were associated with lower macular pigment and was the strongest nondietary predictor of MPOD in our sample, explaining 3% variation in the fully adjusted model. Similar findings have been found in smaller studies, independent of body fat assessment methods (25, 37). Higher body fat was also related to lower serum lutein and zeaxanthin concentrations in our sample, similar to what has been observed by others (37-42). We also observed, similar to Gruber et al (28), that high abdominal body fat was associated with low estimated serum response to dietary lutein and zeaxanthin (Table 3).
Body fat is a major storage site for carotenoids (43, 44), and the sheer mass of larger body fat compartments may shift the distribution of carotenoids away from the blood and retina. We speculate that higher concentrations of carotenoids in body fat of females of reproductive age may confer evolutionary advantage by influencing the carotenoid content of breast milk.
Having diabetes was also an important determinant of lower MPOD in the CAREDS sample even after adjustment for other measured correlates. This is consistent with previous observations of lower levels of MPOD in 26 diabetic patients than in 30 patients without diabetes (45). Lower MPOD in women with diabetes does not reflect lower current or past carotenoid intake. In the present sample, women with diabetes had similar average intakes of lutein and zeaxanthin compared with women without diabetes (2.3 compared with 2.4 mg/d at the time of CAREDS and 2.4 compared with 2.1 mg/d approximately 15 y before CAREDS, respectively). It is possible that absorption of lutein and zeaxanthin is lower in persons with diabetes than in persons without diabetes. Concentrations of lutein and zeaxanthin in the serum were lower in women with diabetes in this sample than in women without diabetes after adjustment for diet intake (though this was not statistically significant after adjustment for waist circumference), and having diabetes was more common in women estimated to be low responders to dietary lutein and zeaxanthin than in women who were high responders (Table 3). However, in a small study of lutein absorption in persons with diabetes (n = 7) and in control participants (n = 5), lutein absorption did not differ (46). It is possible that lower retinal concentrations of carotenoids in persons with diabetes reflect a higher turnover of carotenoids as a result of oxidative stress. Persons with diabetes have more oxidative damage to DNA than do those without diabetes (47), possibly because hyperglycemia generates reactive oxygen species and attenuates antioxidant mechanisms (48).
Alternatively, lower MPOD in women with higher abdominal obesity and diabetes may indicate that these conditions are markers for lipoprotein distributions that reduce the uptake of lutein and zeaxanthin into the blood and tissues. One mechanism by which abdominal fat and diabetes may lower macular pigment density is by influencing the lipoproteins that carry carotenoids to increase the uptake of lutein and zeaxanthin in the retina. These conditions are associated with atherogenic lipoprotein profiles that are characterized by lower concentrations of HDL and higher concentrations of blood triacylglycerols and LDL. We also observed that other characteristics that are often associated with having more atherogenic lipid profiles (ie, higher serum triacylglycerols and dietary saturated fats and lower reported levels of physical activity and alcohol intake) were more common in women who had low serum concentrations of lutein and zeaxanthin despite consuming average levels in the diet.
More atherogenic lipoproteins have lower concentrations of lutein and zeaxanthin. Compared with other carotenoids, a larger proportion of lutein and zeaxanthin are carried in HDL particles (49, 50), and small, dense LDL particles have lower lutein and zeaxanthin than do other lipoproteins (51). Thus, having more abundant small LDL particles or low HDL particles may be related to a lower transport of lutein and zeaxanthin into retinal tissue. The possibility that HDL-cholesterol concentrations influence carotenoid distribution is also supported by evidence that Wisconsin hypoalpha mutant chickens, which are unable to produce stable HDL particles, do not accumulate xanthophyll carotenoids in blood and tissues (52).
Other dietary factors
Fiber, fat, and other carotenoids are known to influence the absorption of lutein and zeaxanthin [reviewed in Zaripheh and Erdman (53)]. In the present sample, the intake of polyunsaturated fats was directly related to MPOD. A common source of polyunsaturated fats in the diet is salad oil. Its relation to retinal carotenoids could also reflect the higher bioavailability of lutein from salads if eaten with dressings that contain fat. The intestinal uptake of carotenoids from salads was shown to be improved with the addition of full-fat salad dressings (54) or avocado (55).
The intake of fiber was also directly related to MPOD in the present sample. Fiber and fruit intake were also directly related to MPOD in 280 men and women aged between 18 and 50 y living in the Midwest (21). The relation of fiber to retinal carotenoids could reflect the higher intake of natural antioxidants from whole grains, fruits, and vegetables, for which fiber is a source.
Iris color, age, and smoking
We observed no significant relation of iris color to MPOD before or after adjustment for other predictors. In 3 previous samples of 95 to 280 participants, those with dark brown irises had higher MPOD than did those with blue or gray irises (21, 23, 56). No relation to iris color was observed in one prior sample (17), and a relation only in men was observed in another (13).
We also observed no significant relation of MPOD to age, which is consistent with observations in some other studies (13, 21). However, the CAREDS sample was limited to persons aged 55 y. Lower MPOD with increasing age was observed in some previous samples that included young and middle-aged adults (17, 23, 37).
Cigarette smoking was unrelated to MPOD in the CAREDS sample, which is consistent with observations in several samples (13, 17, 21) but not in 2 other previous samples (23, 24). In the latter studies, the inverse association with macular pigment was stronger in heavy, compared with light, smokers (23). The low proportion of CAREDS participants who smoked (4%) may have limited our ability to detect lower MPOD rates in smokers, if they exist.
Sex
Previous studies have more consistently observed lower macular pigment in women than in men (13, 23, 57) or similar MPOD in both sexes despite higher intake of carotenoids in women (21). Evidence also exists for sex-related differences in the accumulation of carotenoids in adipose tissue and the retina. Higher adipose tissue concentrations of lutein and zeaxanthin have been observed in women than in men, despite lower concentrations of these carotenoids in the retina of women (7, 13). Moreover, concentrations of these carotenoids in adipose tissue were directly associated with macular pigment in men but not in women in these 2 samples. In quail, lutein accumulation in the retina in response to supplementation was inversely related to lutein accumulation in adipose in females, but directly related to adipose lutein in males (58). Taken together, this leads to the speculation that males concentrate carotenoids in the retina, whereas women preferentially accumulate carotenoids in adipose. This would have an evolutionary advantage, because the ability of women to concentrate carotenoids in adipose may increase the carotenoid concentrations in breast milk. Lutein and zeaxanthin have been observed to represent 25% of total carotenoids in breast milk 4 d postpartum and 50% at 32 d (59). Additional exploration of possible sex-related differences in the accumulation of carotenoids in the retina in response to supplementation is needed.
Unexplained variability and unidentified influences
In the CAREDS sample, dietary lutein and zeaxanthin explained 2% of the variability in MPOD, whereas concentrations of these carotenoids in serum (collected 5–7 y earlier) explained 5% of the variability after adjustment for other measured correlates. Lutein and zeaxanthin in serum samples, which were taken at the time of retinal measurements, would probably have been more strongly correlated with MPOD. The stronger correlations observed with serum concentrations of lutein and zeaxanthin than with dietary lutein and zeaxanthin are likely to reflect, in part, the presence of dietary or endogenous influences on the amount of carotenoids absorbed and on their turnover in blood. One such influence that was not assessed was the presence of inflammation. In the third National Health and Nutrition Examination Survey, high concentrations of C-reactive protein were associated with low concentrations of serum lutein and zeaxanthin in older adults (29, 60). Weaker relations with dietary intake could also partially reflect errors in measurement of dietary levels of lutein and zeaxanthin, such as those that can be attributed to the variations in the carotenoid content of foods, which can be due to different growing locations and varieties of foods. The lower variability in MPOD observed with dietary rather than serum concentrations of lutein and zeaxanthin may also reflect differences in the absorption of carotenoids as a result of cooking or processing or because of the fat content of the specific meal (61). Unexplained variability in MPOD due to errors in the in vivo measurements is also likely.
We were unable to account for the lower average MPOD levels observed in Wisconsin women than in women in Iowa or Oregon with the variables we measured. The lower levels of MPOD observed in Wisconsin women persisted across the women after stratification by all subgroups of factors related to MPOD in this sample [ie, quartile of waist circumference, above and below median serum concentrations of lutein and zeaxanthin, and women with and without diabetes (Mares, unpublished observations)]. It is possible that differences across study sites reflect differences in the execution of the protocol that measures macular pigment or the instrumentation, but this is unlikely. The measurement protocol was highly standardized across sites, and MPOD levels of one study site, which were measured by different examiners and instruments, were nearly identical in the right and left eyes across all 3 study sites (data not shown). Moreover, there were no trends for changing levels of MPOD by month of study visit in quality control analyses during the study period. The women from Wisconsin were not, on average, signficantly different from the women from Iowa or Oregon in any of the attributes measured (data not shown). Thus, lower MPOD levels in the Wisconsin women most likely reflect unknown and unmeasured predictors of MPOD.
Because of the low amount of variability (12%) that can be explained by the numerous dietary, health, and lifestyle factors assessed in CAREDS and because of the inability to account for lower MPOD in the Wisconsin women in this sample, there appears to be one or more unknown important predictors of MPOD. This fact, together with the highly specific nature of the accumulation of these precise carotenoid isomers in the retina, suggests that carrier proteins may play a role in the magnitude of uptake into the retina (62). Results of a recent study in twins suggest that 70–80% of macular pigment density levels are inherited, but this may reflect a combination of influences on carotenoid uptake in the gastrointestinal tract, distribution of carotenoids in the blood and eye and metabolism (63). Research that describes the distribution of carotenoid binding proteins across people and the factors which modulate them may provide additional insights into the variations in the accumulation of carotenoids in eye tissues.
Conclusions
The present study described the nature of the relation of dietary lutein and zeaxanthin with concentrations of these carotenoids in the retina in a sample that is far larger than those of previous studies and which used the first standardized protocol for MPOD measurement. The results confirmed obesity as an important determinant of retinal carotenoids and indicate that diabetes is an important predictor of low retinal carotenoids. Future studies are needed to measure the influence of these attributes on the ability to accumulate carotenoids in the retina. Other important unknown predictors of retinal carotenoids remain to be identified.
|
ACKNOWLEDGMENTS |
|---|
We thank the CAREDS Scientific Advisory Board for their useful discussions and critical reading of the manuscript: Natalie Kurinji, Sheila West, Neil Bressler, Anne Lindblad, and Susan Mayne. We also thank the CAREDS staff for their dedicated work: Paula Smith, Kelly O'Berry, Heather Stockman, Steven Wallace, Lindsey Fuhrmeister, Jane Armstrong, Michael Gruber, Larry Hubbard, Niyati Mehta, Amy Millen-Brady, Michael Neider, Hugh Wabers, Janet Rowley, Tanya Judge, Lisa Oxton, Rick Voland, Gail Ostrowski, Scott Burfield, Tara LaRowe, Julie Ewing, and Tracy Perkins.
We also thank and acknowledge the Women's Health Initiative investigators, staff, and participants for their time and effort in obtaining the WHI data that were presented in this manuscript. Specifically, we would like to thank Barbara Alving, Jacques Rossouw, Linda Pottern, Ross Prentice, Garnet Anderson, Andrea LaCroix, Charles L Kooperberg, Ruth E Patterson, Anne McTiernan, Sally Shumaker, Evan Stein, Steven Cummings, Sylvia Wassertheil-Smoller, Jennifer Hays, JoAnn Manson, Annlouise R Assaf, Lawrence Phillips, Shirley Beresford, Judith Hsia, Rowan Chlebowski, Evelyn Whitlock, Bette Caan, Jane Morley Kotchen, Barbara V Howard, Linda Van Horn, Henry Black, Marcia L Stefanick, Dorothy Lane, Rebecca Jackson, Cora E Lewis, Tamsen Bassford, Jean Wactawski-Wende, John Robbins, Allan Hubbell, Howard Judd, Robert D Langer, Margery Gass, Marian Limacher, David Curb, Robert Wallace, Judith Ockene, Norman Lasser, Mary Jo O'Sullivan, Karen Margolis, Robert Brunner, Gerardo Heiss, Lewis Kuller, Karen C Johnson, Robert Brzyski, Gloria E Sarto, Denise Bonds, and Susan Hendrix.
The CAREDS Macular Pigment Study Group are Barbara Blodi, Alvin Eisner, Karen Gehrs, Randy Kardon, Michael L Klein, Julie A Mares, D Max Snodderly, James Ver Hoeve, and Billy R Wooten.
The CAREDS Investigators are Catherine Allen, Barbara Blodi, Matthew Davis, Robert Wallace, Karen Gehrs, Michael Klein, Cheryl Ritenbaugh, D Max Snodderly, and Elizabeth Johnson.
JAM and all CAREDS Investigators contributed to the design of the study and collection of the data, TLL and MJG also contributed to the collection of data. JAM, TLL, DMS, SMM, MJG, and RJC contributed to the analysis and interpretation of data. EJJ oversaw the analysis of blood carotenoids. JAM, TLL, DMS, SMM, MJG, MLK, BRW and EJJ contributed to preparation of the manuscript. None of the authors had conflicts of interest.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONSUBJECTS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. M. Moeller, R. Voland, G. E. Sarto, V. L. Gobel, S. L. Streicher, and J. A. Mares Women's Health Initiative Diet Intervention Did Not Increase Macular Pigment Optical Density in an Ancillary Study of a Subsample of the Women's Health Initiative J. Nutr., September 1, 2009; 139(9): 1692 - 1699. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. F. Spaide Potential Causes of Altered Autofluorescence in Diabetic Persons Arch Ophthalmol, July 1, 2009; 127(7): 942 - 943. [Full Text] [PDF]
|
|
|
|
 |
|
 L. Zhao and B. V. Sweet Lutein and zeaxanthin for macular degeneration Am. J. Health Syst. Pharm., July 1, 2008; 65(13): 1232 - 1238. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Nolan, J. M. Stringham, S. Beatty, and D. M. Snodderly Spatial Profile of Macular Pigment and Its Relationship to Foveal Architecture Invest. Ophthalmol. Vis. Sci., May 1, 2008; 49(5): 2134 - 2142. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Beatty, F. J. G. M. van Kuijk, and U. Chakravarthy Macular Pigment and Age-Related Macular Degeneration: Longitudinal Data and Better Techniques of Measurement Are Needed Invest. Ophthalmol. Vis. Sci., March 1, 2008; 49(3): 843 - 845. [Full Text] [PDF]
|
|
|
|
|
|
S. M. Moeller, R. Voland, L. Tinker, B. A. Blodi, M. L. Klein, K. M. Gehrs, E. J. Johnson, D. M. Snodderly, R. B. Wallace, R. J. Chappell, et al. Associations Between Age-Related Nuclear Cataract and Lutein and Zeaxanthin in the Diet and Serum in the Carotenoids in the Age-Related Eye Disease Study (CAREDS), an Ancillary Study of the Women's Health Initiative Arch Ophthalmol, March 1, 2008; 126(3): 354 - 364. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. E. Connor, P. B. Duell, R. Kean, and Y. Wang The Prime Role of HDL to Transport Lutein into the Retina: Evidence from HDL-Deficient WHAM Chicks Having a Mutant ABCA1 Transporter Invest. Ophthalmol. Vis. Sci., September 1, 2007; 48(9): 4226 - 4231. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, Right eye high lutein; 
, left eye high lutein; •, right eye low lutein; 
, left eye low lutein; solid line, exponentiated curve right eye high lutein (y = 0.6807e–0.9251x); dashed line, exponentiated curve right eye low lutein (y = 0.4937e–0.9007x).
) and after (
; n = 1590) diabetes (P for interaction = 0.05) after adjustment for waist circumference, study visit site, dietary fiber, and polyunsaturated fats (% of energy; obtained by multiple linear regression).